arXiv subjects

Ziyao Wang

Publications and source records attributed to Ziyao Wang.

At least 55 records · Page 3Linked to original sources

arXiv2025

Lévy-Driven Option Pricing without a Riskless Asset

We extend the Lindquist-Rachev (LR) option-pricing framework--which values derivatives in markets lacking a traded risk-free bond--by introducing common Levy jump dynamics across two risky assets. The resulting endogenous "shadow" short rate replaces the usual risk-free yield and governs discounting and risk-neutral drifts. We focus on two widely used pure-jump specifications: the Normal Inverse Gaussian (NIG) process and the Carr-Geman-Madan-Yor (CGMY) tempered-stable process. Using Ito-Levy calculus we derive an LR partial integro-differential equation (LR-PIDE) and obtain European option values through characteristic-function methods implemented with the Fast Fourier Transform (FFT) and Fourier-cosine (COS) algorithms. Calibrations to S and P 500 index options show that both jump models materially reduce pricing errors and fit the observed volatility smile far better than the Black-Scholes benchmark; CGMY delivers the largest improvement. We also extract time-varying shadow short rates from paired asset data and show that sharp declines coincide with liquidity-stress episodes, highlighting risk signals not visible in Treasury yields. The framework links jump risk, relative asset pricing, and funding conditions in a tractable form for practitioners.

q-fin.MF↗

arXiv2025

EdgeLoRA: An Efficient Multi-Tenant LLM Serving System on Edge Devices

Large Language Models (LLMs) have gained significant attention due to their versatility across a wide array of applications. Fine-tuning LLMs with parameter-efficient adapters, such as Low-Rank Adaptation (LoRA), enables these models to efficiently adapt to downstream tasks without extensive retraining. Deploying fine-tuned LLMs on multi-tenant edge devices offers substantial benefits, such as reduced latency, enhanced privacy, and personalized responses. However, serving LLMs efficiently on resource-constrained edge devices presents critical challenges, including the complexity of adapter selection for different tasks and memory overhead from frequent adapter swapping. Moreover, given the multiple requests in multi-tenant settings, processing requests sequentially results in underutilization of computational resources and increased latency. This paper introduces EdgeLoRA, an efficient system for serving LLMs on edge devices in multi-tenant environments. EdgeLoRA incorporates three key innovations: (1) an adaptive adapter selection mechanism to streamline the adapter configuration process; (2) heterogeneous memory management, leveraging intelligent adapter caching and pooling to mitigate memory operation overhead; and (3) batch LoRA inference, enabling efficient batch processing to significantly reduce computational latency. Comprehensive evaluations using the Llama3.1-8B model demonstrate that EdgeLoRA significantly outperforms the status quo (i.e., llama.cpp) in terms of both latency and throughput. The results demonstrate that EdgeLoRA can achieve up to a 4 times boost in throughput. Even more impressively, it can serve several orders of magnitude more adapters simultaneously. These results highlight EdgeLoRA's potential to transform edge deployment of LLMs in multi-tenant scenarios, offering a scalable and efficient solution for resource-constrained environments.

cs.DC↗

arXiv2025

EvoVerilog: Large Langugage Model Assisted Evolution of Verilog Code

Large Language Models (LLMs) have demonstrated great potential in automating the generation of Verilog hardware description language code for hardware design. This automation is critical to reducing human effort in the complex and error-prone process of hardware design. However, existing approaches predominantly rely on human intervention and fine-tuning using curated datasets, limiting their scalability in automated design workflows. Although recent iterative search techniques have emerged, they often fail to explore diverse design solutions and may underperform simpler approaches such as repeated prompting. To address these limitations, we introduce EvoVerilog, a novel framework that combines the reasoning capabilities of LLMs with evolutionary algorithms to automatically generate and refine Verilog code. EvoVerilog utilizes a multiobjective, population-based search strategy to explore a wide range of design possibilities without requiring human intervention. Extensive experiments demonstrate that EvoVerilog achieves state-of-the-art performance, with pass@10 scores of 89.1 and 80.2 on the VerilogEval-Machine and VerilogEval-Human benchmarks, respectively. Furthermore, the framework showcases its ability to explore diverse designs by simultaneously generating a variety of functional Verilog code while optimizing resource utilization.

cs.AR↗

arXiv2025

Invisible Tokens, Visible Bills: The Urgent Need to Audit Hidden Operations in Opaque LLM Services

Modern large language model (LLM) services increasingly rely on complex, often abstract operations, such as multi-step reasoning and multi-agent collaboration, to generate high-quality outputs. While users are billed based on token consumption and API usage, these internal steps are typically not visible. We refer to such systems as Commercial Opaque LLM Services (COLS). This position paper highlights emerging accountability challenges in COLS: users are billed for operations they cannot observe, verify, or contest. We formalize two key risks: \textit{quantity inflation}, where token and call counts may be artificially inflated, and \textit{quality downgrade}, where providers might quietly substitute lower-cost models or tools. Addressing these risks requires a diverse set of auditing strategies, including commitment-based, predictive, behavioral, and signature-based methods. We further explore the potential of complementary mechanisms such as watermarking and trusted execution environments to enhance verifiability without compromising provider confidentiality. We also propose a modular three-layer auditing framework for COLS and users that enables trustworthy verification across execution, secure logging, and user-facing auditability without exposing proprietary internals. Our aim is to encourage further research and policy development toward transparency, auditability, and accountability in commercial LLM services.

cs.CR↗

arXiv2025

CoIn: Counting the Invisible Reasoning Tokens in Commercial Opaque LLM APIs

As post-training techniques evolve, large language models (LLMs) are increasingly augmented with structured multi-step reasoning abilities, often optimized through reinforcement learning. These reasoning-enhanced models outperform standard LLMs on complex tasks and now underpin many commercial LLM APIs. However, to protect proprietary behavior and reduce verbosity, providers typically conceal the reasoning traces while returning only the final answer. This opacity introduces a critical transparency gap: users are billed for invisible reasoning tokens, which often account for the majority of the cost, yet have no means to verify their authenticity. This opens the door to token count inflation, where providers may overreport token usage or inject synthetic, low-effort tokens to inflate charges. To address this issue, we propose CoIn, a verification framework that audits both the quantity and semantic validity of hidden tokens. CoIn constructs a verifiable hash tree from token embedding fingerprints to check token counts, and uses embedding-based relevance matching to detect fabricated reasoning content. Experiments demonstrate that CoIn, when deployed as a trusted third-party auditor, can effectively detect token count inflation with a success rate reaching up to 94.7%, showing the strong ability to restore billing transparency in opaque LLM services. The dataset and code are available at https://github.com/CASE-Lab-UMD/LLM-Auditing-CoIn.

cs.AI↗

arXiv2025

Three-dimensional topological disclination in acoustic crystals

Topological disclinations, crystallographic defects that break rotation lattice symmetry, have attracted great interest and exhibited wide applications in cavities, waveguides, and lasers. However, topological disclinations have thus far been predominantly restricted to two-dimensional (2D) systems owing to the substantial challenges in constructing such defects in three-dimensional (3D) systems and characterizing their topological features. Here we report the theoretical proposal and experimental demonstration of a 3D topological disclination that exhibits fractional (1/2) charge and zero-dimensional (0D) topological bound states, realized by cutting-and-gluing a 3D acoustic topological crystalline insulator. Using acoustic pump-probe measurements, we directly observe 0D topological disclination states at the disclination core, consistent with the tight-binding model and full-wave simulation results. Our results extend the research frontier of topological disclinations and open a new paradigm for exploring the interplay between momentum-space band topology and the real-space defect topology in 3D and higher dimensions.

physics.app-ph↗

arXiv2025

FedHQ: Hybrid Runtime Quantization for Federated Learning

Federated Learning (FL) is a decentralized model training approach that preserves data privacy but struggles with low efficiency. Quantization, a powerful training optimization technique, has been widely explored for integration into FL. However, many studies fail to consider the distinct performance attribution between particular quantization strategies, such as post-training quantization (PTQ) or quantization-aware training (QAT). As a result, existing FL quantization methods rely solely on either PTQ or QAT, optimizing for speed or accuracy while compromising the other. To efficiently accelerate FL and maintain distributed convergence accuracy across various FL settings, this paper proposes a hybrid quantitation approach combining PTQ and QAT for FL systems. We conduct case studies to validate the effectiveness of using hybrid quantization in FL. To solve the difficulty of modeling speed and accuracy caused by device and data heterogeneity, we propose a hardware-related analysis and data-distribution-related analysis to help identify the trade-off boundaries for strategy selection. Based on these, we proposed a novel framework named FedHQ to automatically adopt optimal hybrid strategy allocation for FL systems. Specifically, FedHQ develops a coarse-grained global initialization and fine-grained ML-based adjustment to ensure efficiency and robustness. Experiments show that FedHQ achieves up to 2.47x times training acceleration and up to 11.15% accuracy improvement and negligible extra overhead.

cs.LG↗

arXiv2025

Speculate, then Collaborate: Fusing Knowledge of Language Models during Decoding

Large Language Models (LLMs) often excel in specific domains but fall short in others due to the limitations of their training. Thus, enabling LLMs to solve problems collaboratively by integrating their complementary knowledge promises to improve their performance across domains. To realize this potential, we introduce a novel Collaborative Speculative Decoding (CoSD) algorithm that enables efficient LLM knowledge fusion at test time without requiring additional model training. CoSD employs a draft model to generate initial sequences and an easy-to-learn rule or decision tree to decide when to invoke an assistant model to improve these drafts. CoSD not only enhances knowledge fusion but also improves inference efficiency, is transferable across domains and models, and offers greater explainability. Experimental results demonstrate that CoSD improves accuracy by up to 10\% across benchmarks compared to existing methods, providing a scalable and effective solution for LLM-based applications

cs.CL↗

arXiv2025

Prada: Black-Box LLM Adaptation with Private Data on Resource-Constrained Devices

In recent years, Large Language Models (LLMs) have demonstrated remarkable abilities in various natural language processing tasks. However, adapting these models to specialized domains using private datasets stored on resource-constrained edge devices, such as smartphones and personal computers, remains challenging due to significant privacy concerns and limited computational resources. Existing model adaptation methods either compromise data privacy by requiring data transmission or jeopardize model privacy by exposing proprietary LLM parameters. To address these challenges, we propose Prada, a novel privacy-preserving and efficient black-box LLM adaptation system using private on-device datasets. Prada employs a lightweight proxy model fine-tuned with Low-Rank Adaptation (LoRA) locally on user devices. During inference, Prada leverages the logits offset, i.e., difference in outputs between the base and adapted proxy models, to iteratively refine outputs from a remote black-box LLM. This offset-based adaptation approach preserves both data privacy and model privacy, as there is no need to share sensitive data or proprietary model parameters. Furthermore, we incorporate speculative decoding to further speed up the inference process of Prada, making the system practically deployable on bandwidth-constrained edge devices, enabling a more practical deployment of Prada. Extensive experiments on various downstream tasks demonstrate that Prada achieves performance comparable to centralized fine-tuning methods while significantly reducing computational overhead by up to 60% and communication costs by up to 80%.

cs.CR↗

arXiv2025

NeuroSymAD: A Neuro-Symbolic Framework for Interpretable Alzheimer's Disease Diagnosis

Alzheimer's disease (AD) diagnosis is complex, requiring the integration of imaging and clinical data for accurate assessment. While deep learning has shown promise in brain MRI analysis, it often functions as a black box, limiting interpretability and lacking mechanisms to effectively integrate critical clinical data such as biomarkers, medical history, and demographic information. To bridge this gap, we propose NeuroSymAD, a neuro-symbolic framework that synergizes neural networks with symbolic reasoning. A neural network percepts brain MRI scans, while a large language model (LLM) distills medical rules to guide a symbolic system in reasoning over biomarkers and medical history. This structured integration enhances both diagnostic accuracy and explainability. Experiments on the ADNI dataset demonstrate that NeuroSymAD outperforms state-of-the-art methods by up to 2.91% in accuracy and 3.43% in F1-score while providing transparent and interpretable diagnosis.

eess.IV↗

arXiv2025

Towards counterfactual fairness through auxiliary variables

The challenge of balancing fairness and predictive accuracy in machine learning models, especially when sensitive attributes such as race, gender, or age are considered, has motivated substantial research in recent years. Counterfactual fairness ensures that predictions remain consistent across counterfactual variations of sensitive attributes, which is a crucial concept in addressing societal biases. However, existing counterfactual fairness approaches usually overlook intrinsic information about sensitive features, limiting their ability to achieve fairness while simultaneously maintaining performance. To tackle this challenge, we introduce EXOgenous Causal reasoning (EXOC), a novel causal reasoning framework motivated by exogenous variables. It leverages auxiliary variables to uncover intrinsic properties that give rise to sensitive attributes. Our framework explicitly defines an auxiliary node and a control node that contribute to counterfactual fairness and control the information flow within the model. Our evaluation, conducted on synthetic and real-world datasets, validates EXOC's superiority, showing that it outperforms state-of-the-art approaches in achieving counterfactual fairness. Our code is available at https://github.com/CASE-Lab-UMD/counterfactual_fairness_2025.

cs.LG↗

arXiv2025

Kolmogorov-Arnold Network for Remote Sensing Image Semantic Segmentation

Semantic segmentation plays a crucial role in remote sensing applications, where the accurate extraction and representation of features are essential for high-quality results. Despite the widespread use of encoder-decoder architectures, existing methods often struggle with fully utilizing the high-dimensional features extracted by the encoder and efficiently recovering detailed information during decoding. To address these problems, we propose a novel semantic segmentation network, namely DeepKANSeg, including two key innovations based on the emerging Kolmogorov Arnold Network (KAN). Notably, the advantage of KAN lies in its ability to decompose high-dimensional complex functions into univariate transformations, enabling efficient and flexible representation of intricate relationships in data. First, we introduce a KAN-based deep feature refinement module, namely DeepKAN to effectively capture complex spatial and rich semantic relationships from high-dimensional features. Second, we replace the traditional multi-layer perceptron (MLP) layers in the global-local combined decoder with KAN-based linear layers, namely GLKAN. This module enhances the decoder's ability to capture fine-grained details during decoding. To evaluate the effectiveness of the proposed method, experiments are conducted on two well-known fine-resolution remote sensing benchmark datasets, namely ISPRS Vaihingen and ISPRS Potsdam. The results demonstrate that the KAN-enhanced segmentation model achieves superior performance in terms of accuracy compared to state-of-the-art methods. They highlight the potential of KANs as a powerful alternative to traditional architectures in semantic segmentation tasks. Moreover, the explicit univariate decomposition provides improved interpretability, which is particularly beneficial for applications requiring explainable learning in remote sensing.

cs.CV↗

arXiv2024

Topological Dirac-vortex modes in a three-dimensional photonic topological insulator

Recently, topological Dirac-vortex modes in Kekulé-distorted photonic lattices have attracted broad interest and exhibited promising applications in robust photonic devices such as topological cavities, lasers, and fibers. However, due to the vectorial nature of electromagnetic waves that results in complicated band dispersions and fails the tight-binding model predictions, it is challenging to construct three-dimensional (3D) topological photonic structures with Kekulé distortion and the photonic topological Dirac-vortex modes have thus far been limited to two-dimensional (2D) systems. Here, by directly mapping a 3D Kekulé-distorted tight-binding model in a 3D tight-binding-like photonic crystal exhibiting scalar-wave-like band structures, we theoretically propose and experimentally demonstrate topological Dirac-vortex modes in a 3D photonic topological insulator for the first time. Using microwave near-field measurements, we directly observe robust photonic topological Dirac-vortex modes bound to and propagate along a one-dimensional (1D) Dirac-vortex line defect, matching well with the tight-binding and simulation results. Our work offers an ideal platform to map tight-binding models in 3D topological photonic crystals directly and opens a new avenue for exploiting topological lattice defects to manipulate light in 3D space.

physics.optics↗

arXiv2024

SHED: Shapley-Based Automated Dataset Refinement for Instruction Fine-Tuning

The pre-trained Large Language Models (LLMs) can be adapted for many downstream tasks and tailored to align with human preferences through fine-tuning. Recent studies have discovered that LLMs can achieve desirable performance with only a small amount of high-quality data, suggesting that a large amount of the data in these extensive datasets is redundant or even harmful. Identifying high-quality data from vast datasets to curate small yet effective datasets has emerged as a critical challenge. In this paper, we introduce SHED, an automated dataset refinement framework based on Shapley value for instruction fine-tuning. SHED eliminates the need for human intervention or the use of commercial LLMs. Moreover, the datasets curated through SHED exhibit transferability, indicating they can be reused across different LLMs with consistently high performance. We conduct extensive experiments to evaluate the datasets curated by SHED. The results demonstrate SHED's superiority over state-of-the-art methods across various tasks and LLMs; notably, datasets comprising only 10% of the original data selected by SHED achieve performance comparable to or surpassing that of the full datasets.

cs.CL↗

arXiv2024

Moderator: Moderating Text-to-Image Diffusion Models through Fine-grained Context-based Policies

We present Moderator, a policy-based model management system that allows administrators to specify fine-grained content moderation policies and modify the weights of a text-to-image (TTI) model to make it significantly more challenging for users to produce images that violate the policies. In contrast to existing general-purpose model editing techniques, which unlearn concepts without considering the associated contexts, Moderator allows admins to specify what content should be moderated, under which context, how it should be moderated, and why moderation is necessary. Given a set of policies, Moderator first prompts the original model to generate images that need to be moderated, then uses these self-generated images to reverse fine-tune the model to compute task vectors for moderation and finally negates the original model with the task vectors to decrease its performance in generating moderated content. We evaluated Moderator with 14 participants to play the role of admins and found they could quickly learn and author policies to pass unit tests in approximately 2.29 policy iterations. Our experiment with 32 stable diffusion users suggested that Moderator can prevent 65% of users from generating moderated content under 15 attempts and require the remaining users an average of 8.3 times more attempts to generate undesired content.

cs.CR↗

arXiv2024

FLoRA: Federated Fine-Tuning Large Language Models with Heterogeneous Low-Rank Adaptations

The rapid development of Large Language Models (LLMs) has been pivotal in advancing AI, with pre-trained LLMs being adaptable to diverse downstream tasks through fine-tuning. Federated learning (FL) further enhances fine-tuning in a privacy-aware manner by utilizing clients' local data through in-situ computation, eliminating the need for data movement. However, fine-tuning LLMs, given their massive scale of parameters, poses challenges for clients with constrained and heterogeneous resources in FL. Previous methods employed low-rank adaptation (LoRA) for efficient federated fine-tuning but utilized traditional FL aggregation strategies on LoRA adapters. These approaches led to mathematically inaccurate aggregation noise, reducing fine-tuning effectiveness and failing to address heterogeneous LoRAs. In this work, we first highlight the mathematical incorrectness of LoRA aggregation in existing federated fine-tuning methods. We introduce a new approach called FLORA that enables federated fine-tuning on heterogeneous LoRA adapters across clients through a novel stacking-based aggregation method. Our approach is noise-free and seamlessly supports heterogeneous LoRA adapters. Extensive experiments demonstrate FLORA' s superior performance in both homogeneous and heterogeneous settings, surpassing state-of-the-art methods. We envision this work as a milestone for efficient, privacy-preserving, and accurate federated fine-tuning of LLMs. Our code is available at https://github.com/ATP-1010/FederatedLLM.

cs.LG↗

arXiv2024

Realization of type-II double-zero-index photonic crystals

Some photonic crystals (PCs) with Dirac-like conical dispersions exhibit the property of double zero refractive index (that is, both epsilon and mu near zero (EMNZ)), wherein the electromagnetic waves have an infinite effective wavelength and do not experience any spatial phase change. The Dirac-like cones that support EMNZ are previously thought to present only at the center of the Brillouin zone ($Γ$ point) with a zero wavevector (we refer to as type-I EMNZ), which is constrained by the proportional relationship between phase refractive index and wavevector ($n=kc/ω$). Here, we demonstrate the existence of an anomalous type-II EMNZ in PCs, which is associated with the Dirac-like point at off-$Γ$ points. By introducing a wave modulation approach, we theoretically elucidate its physical mechanism, and resolve the paradox of type-II EMNZ with non-zero wavevectors. We then fabricate a type-II EMNZ PC operating at the X point, and experimentally demonstrate that both its effective permittivity and permeability are zero at the Dirac-like point. Type-II EMNZ PCs exhibit a range of intriguing phenomena, including angle-selective transmission, wavefront flattening, a 180$^{\circ}$ phase shift upon transmission, and waveguiding with natural zero radiation loss. The extraordinary properties of type-II EMNZ PCs may open new avenues for the development of angle-selective optical filters, directional light sources, phase-controlled optical switches, ultracompact photonic circuits, nanolasers, and on-chip nonlinear enhancement.

physics.optics↗

arXiv2024

Photonic bilayer Chern insulator with corner states

Photonic Chern insulators can be implemented in gyromagnetic photonic crystals with broken time-reversal (TR) symmetry. They exhibit gapless chiral edge states (CESs), enabling unidirectional propagation and demonstrating exceptional resilience to localization even in the presence of defects or disorders. However, when two Chern insulators with opposite Chern numbers are stacked together, this one-way nature can be nullified, causing the originally gapless CESs to become gapped. Recent theoretical works have proposed achieving such a topological phase transition in condensed matter systems using antiferromagnetic thin films such as MnBi2Te4 or by coupling two quantum spin/anomalous Hall insulators, but these approaches have yet to be realized experimentally. In a bilayer gyromagnetic photonic crystal arranged in an antiferromagnetic layer configuration, our experimental observations reveal that interlayer coupling initiates a transition from a Chern insulating phase to a higher-order topological phase. This transition results in the gapping of CESs and triggers the emergence of corner states within the bandgap. The corner mode energy within the gap can be attributed to CESs interaction, forming a Jackiw-Rebbi topological domain wall mode at the corner. These states exhibit heightened resilience against defects, setting them apart from their time-reversal symmetric counterparts.

cond-mat.mes-hall↗